Empirical Evaluation of Visual Graph Analytic Techniques

Visual Graph analytics is one of the best sources for creating its remarkable, distinct impact in the field of data science. In most new time graph analytics and big data has fascinated a wide range of attention from the researchers and scientists from all over the world. By using the most advanced tools for the graph, analytics can lead the most useful and productive results in various domains which include life sciences, business, computer sciences, engineering and so on. Biological data can be represented in interpretable form when exposed to graph analytic tools which may lead to meaningful insights. This research paper is aimed for the empirical evaluation of three tools MEME suite, DMINDA, and Neo4j. In this paper, we have various research objectives such as a collection of datasets from heterogeneous biological data sources, data integration, and formation of the new dataset (MYBIOGRID). Design the queries in Neo4j using Cypher Query Language in order to visualize MYBIOGRID using Property Graph Model. Determine the relationships in the dataset, based on path analysis and centrality analysis; uploading data in MEME suite and DMINDA for motif elicitation. The result from this study indicates that visualization of similarity matrix of repetitive patterns thus representing the most similar and least similar patterns in the sequence shown in a conclusion table. Neo4j graph databases play a vital role in graph analytics but in memory storage makes analysis very lime to consume for the massive bulk of data sets. Each tool has its specific parameters making it a good candidate for analysis.